FYI, I got this working…kind of.
After getting past the JNDI issue (just had to set the JNDI name of the
ClusterTopic and JmsConnectionFactory), I had a problem where EclipseLink would
throw an exception upon receipt of a cache coordination message because the
message's destination was null.
To make it work, I had to use a Message Driven Bean, inject an instance of
ClusterTopic in that bean and explicitly set the message destination in the
onMessage method of the MDB (setting it to the injected ClusterTopic instance).
It doesn't seem right that the message destination is null, but if this is a
legitimate work-around then I'm fine with it.
Can anyone shed more light on this? Is this a legitimate work-around?
On Nov 14, 2012, at 11:27 PM, Dan Ziegelbein wrote:
> I am trying to figure out how to get EclipseLink's cache coordination working
> within a Resin cluster. I am trying to get it working using JMS.
> So far I haven't been able to get past the JNDI lookup of the
> JmsConnectionFactory (which is failing).
> Right now my knowledge on this subject is a shaky patchwork of tidbits from a
> bunch of google searches, and I'm hoping that I'm missing some simple
> configuration step. But, before getting into all that...Should I even be able
> to do this?
> Currently using Resin Pro 4.0.23.
> resin-interest mailing list
resin-interest mailing list